Determining the Association between Child Dietary Behaviors and Weight Status

Abstract

IntroductionIn 2011–2012, over a third of U.S. children were overweight or obese and 17% were obese. The Expert Committee on the Assessment, Prevention and Treatment of Child and Adolescent Overweight and Obesity created a list of behaviors with recommendations to prevent or treat overweight and obesity in children. A screening tool was developed at the Rush Pediatric Primary Care Center to start the conversation between physicians and registered dietitian nutritionists with families about these behaviors.ObjectiveTo determine the association between the child's weight status, dietary behaviors, and other demographic variables assessed on the Child Nutrition and Physical Activity (CNPA) screener.DesignA cross sectional design study was conducted using a retrospective chart review method.MethodsDietary behavior and parent perception responses were collected using the CNPA screener. Dietary behaviors included sugar sweetened beverage (SSB), fruit and vegetable intake, eating breakfast, eating out, eating dinner with family, number of meals and snacks consumed, and type and amount of milk consumed. A CNPA was given to parents of children who came to the clinic between 2012–2015. Chi square was used to determine the association between weight status, demographic characteristics and dietary behaviors, as well as the association between accurate parent perception and demographic characteristics. Forward logistic regression analysis was conducted with odds of being obese and severe obese compared to healthy weight.ResultsA total of 2,230 children, 2–18 years old were included in the analysis. A majority of the sample was non‐Hispanic black (59.6%) and healthy weight (61.1%). Based on univariate analysis, 17 variables were entered into the forward logistic regression analysis; age, sex, race, drinking SSBs, eating breakfast, fruit and vegetable, eating out, eating dinner with family, number of meals and snacks consumed, milk type consumed, and parent perception of the child's weight status. The forward wald model was statistically significant χ2 (7, N=1312) = 501.01, p <.001, indicating that the model was able to distinguish between healthy weight children and those who were obese or severe obese. The model explained 31.7% to 46.3% of the variance in weight status and correctly classified 84.3% of cases. Increasing age (1.14 OR (95% CI 1.10–1.18) (p<.001), being non‐Hispanic black (2.33 OR (95% CI 1.34–4.06) (p=.003), consuming 2% milk (2.97 OR (95% CI 1.91–4.62) (p<.001), being Hispanic (3.14 OR (95% CI 1.77–5.57) (p<.001), and consuming non‐fat/1% milk (8.83 OR (95% CI 5.01–15.55) (p<.001) increased the odds of being obese or severe obese. Increasing number of meals (.75 OR (95% CI .59–.95) (p<.001) and correctly perceiving the child's weight status (0.03 OR (95% CI 0.02–0.04) (p<.001) decreased the odds of being obese or severe obese.ConclusionThe odds of being obese or severe obese were higher if the child consumed nonfat/1% milk or 2% milk, and lower as number of meals increased and if the parent correctly perceived their child's weight. The other dietary behaviors had no significant association with the odds of being obese or severe obese in this predominately non‐Hispanic black sample.Support or Funding InformationDepartmental Funding